BUG #17350: GIST TRGM Index is broken when combining with combining INCLUDE with a string function (e.g. lower). - Mailing list pgsql-bugs

From PG Bug reporting form
Subject BUG #17350: GIST TRGM Index is broken when combining with combining INCLUDE with a string function (e.g. lower).
Date
Msg-id 17350-b5bdcf476e5badbb@postgresql.org
Whole thread Raw
Responses Re: BUG #17350: GIST TRGM Index is broken when combining with combining INCLUDE with a string function (e.g. lower).  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-bugs
The following bug has been logged on the website:

Bug reference:      17350
Logged by:          louis jachiet
Email address:      louis.jachiet@telecom-paris.fr
PostgreSQL version: 13.4
Operating system:   Archlinux
Description:

Hello everyone,

I hope this is not a duplicate of a similar bug (I tried to check…). GIST
TRGM Index seems to return
wrong values when populated with a string function (such as lower)  and an
include. If I issue the 
following lines, instead of returning 'foo' and 'bar', postgresql will
return  two empty lines:

    CREATE EXTENSION IF NOT EXISTS pg_trgm ;
    CREATE TABLE t (a VARCHAR(50));
    INSERT INTO t VALUES ('foo') ;
    INSERT INTO t VALUES ('BAR') ;
    CREATE INDEX test_idx ON t USING gist(lower(a) gist_trgm_ops) INCLUDE
(a) ;
    set enable_seqscan=off ;
    SELECT lower(a) FROM t ;

Obviously for this last SELECT, the index is only useful because of the
seqscan=off 
but if you have a large database and issue a command like:

    SELECT 1 FROM t WHERE lower(a) LIKE '%o%' ;

then the output will be also empty because the index will return empty lines
and the recheck condition will fail.
If I use a function different than lower (e.g. upper) the problem persists
but if remove the function or if remove the
include then everything works just fine. It really seems that it is the
combination of include + string function + gist trgm
that makes the SELECT return empty lines. 

Thank you for your time!

Regards


pgsql-bugs by date:

Previous
From: Yura Sokolov
Date:
Subject: Re: BUG #17335: Duplicate result rows in Gather node when "Suppress Append ... that have a single child" applied
Next
From: Reid Thompson
Date:
Subject: Re: BUG #17348: Can never install pgsql on Ubuntu